Messages in this thread | | | Date | Fri, 16 Jan 2015 11:53:02 -0500 | From | Steven Rostedt <> | Subject | Re: [PATCH 1/2 v3] ftrace: don't allow IPMODIFY without proper compiler support |
| |
On Fri, 16 Jan 2015 17:41:58 +0100 (CET) Jiri Kosina <jkosina@suse.cz> wrote:
> On Fri, 16 Jan 2015, Steven Rostedt wrote: > > > Instead, make live kernel patching fail to load if fentry isn't > > supported. IOW, instead of ftrace_ipmodify_supported, have a > > live_kernel_patching_supported that could be based on fentry being used > > or not. > > I can live with that, we are handling non-fentry case ourselves already. > > You expressed that you would accept such patch for ftrace in a past, so I
Right, but then I thought about it more :-)
I had a lot of time thinking about things like this while debugging the dance between function graph and jprobes.
> thought I'll move it out from livepatching to ftrace. But if you don't > want it now, that's not a big issue, and handling the possible fallout > stays the responsibility of ftrace_ops "owner". >
Exactly!
-- Steve
| |